NVIDIA GeForce GTX 1660 Ti vs NVIDIA GeForce GTX 950M Comparison

The NVIDIA GeForce GTX 950M and the NVIDIA GeForce GTX 1660 Ti represent two distinct eras of mobile and desktop graphics, separated by four years of architectural evolution. The data reveals a stark contrast: the GTX 1660 Ti dominates in raw compute throughput, while the GTX 950M surprisingly edges out a win in one specific API workload. This analysis walks through the benchmark results, specification gaps, and architectural differences to determine which GPU suits particular use cases, based solely on the provided facts.

Head-to-Head Benchmarks

The head-to-head comparison presents a fascinating split, with each GPU claiming one victory. In the Geekbench OpenCL test, the GTX 1660 Ti delivers a score of 51,029, dwarfing the GTX 950M’s 9,745. This translates to a deltaPct of -80.9, meaning the GTX 950M is roughly 81% behind the GTX 1660 Ti in this compute-oriented benchmark. The magnitude of this gap is enormous—the GTX 1660 Ti achieves over five times the score of the older chip. OpenCL workloads, which often stress raw shader throughput and memory bandwidth, heavily favor the newer architecture.

However, the Geekbench Vulkan test flips the script. Here, the GTX 950M scores 6,525, while the GTX 1660 Ti trails at 5,827, giving the GTX 950M a 12% advantage. This is a surprising result, given the GTX 1660 Ti’s superior specifications elsewhere. Vulkan’s lower-level API overhead can sometimes expose driver optimization differences or clock behavior quirks, and in this instance, the older Maxwell-based GPU outperforms the Turing-based one. It is notably the GTX 1660 Ti’s average benchmark score across all tests is 7,723, while the GTX 950M averages 8,135—meaning the GTX 950M actually holds a higher aggregate average, despite losing the OpenCL test so decisively.

The wins are split evenly at one apiece, but the context matters. The OpenCL delta is catastrophic for the GTX 950M, while the Vulkan delta is modest. The GTX 1660 Ti’s performance in other tests, such as PassMark G3D (12,871) and PassMark GPU Compute (5,226), reinforces its overall superiority, but those tests were not run on the GTX 950M in this data set. Therefore, the head-to-head data alone paints a mixed picture: the GTX 1660 Ti is the clear compute king, but the GTX 950M holds a niche advantage in Vulkan-specific rendering.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 950M has an average benchmark score of 8,135, which is higher than the GTX 1660 Ti’s 7,723 average. This is despite the GTX 1660 Ti’s massive OpenCL win, suggesting the GTX 950M performs more consistently across the tests it was subjected to.

Q: How large is the performance gap in the Geekbench OpenCL test?

A: The GTX 1660 Ti scores 51,029 versus the GTX 950M’s 9,745, resulting in a deltaPct of -80.9. This means the GTX 950M is over 80% slower in this specific workload, representing a multi-generational leap in compute capability.

Q: Does the GTX 950M win any benchmark against the GTX 1660 Ti?

A: Yes, in the Geekbench Vulkan test, the GTX 950M scores 6,525, which is 12% higher than the GTX 1660 Ti’s 5,827. This is the only head-to-head win for the older GPU.

Q: What is the percentile ranking of each GPU among all GPUs?

A: The GTX 950M ranks in the 42nd percentile, while the GTX 1660 Ti sits just below at the 41st percentile. Despite the GTX 1660 Ti’s higher raw performance, both GPUs occupy similar positions in the broader performance distribution.

Q: What are the nearest rivals for each GPU based on average score?

A: For the GTX 950M, the nearest rival is the AMD Radeon R9 M360 with a deltaPct of 0.1, followed by the NVIDIA GeForce GTX 980 at -0.4. For the GTX 1660 Ti, the closest rival is the AMD Radeon 540 at 0.7, with the AMD Radeon Pro WX 3100 at 1.9.

Q: Which GPU has a higher transistor density?

A: The GTX 1660 Ti has a transistor density of 23.2M per mm², while the GTX 950M has a density of 12.6M per mm². This reflects the newer 12nm process node compared to the older 28nm node.

Where Each One Wins

The GTX 1660 Ti is the overwhelming winner for compute-heavy tasks. Its OpenCL score of 51,029 is not just a marginal improvement—it is a fundamental shift in capability. For applications that leverage general-purpose GPU computing, such as rendering, simulation, or machine learning inference, the GTX 1660 Ti is categorically superior. The PassMark DirectX 11 score of 103 and DirectX 12 score of 51 further indicate strong performance in modern graphics APIs, though those tests lack direct comparison data from the GTX 950M.

The GTX 950M’s single win in Vulkan (6,525 vs 5,827) suggests it may hold an edge in specific Vulkan-based game engines or applications where driver overhead or memory access patterns favor the Maxwell architecture. However, this is a narrow victory. The GTX 950M also holds a higher average benchmark score (8,135 vs 7,723) and a higher percentile ranking (42 vs 41), indicating that across its tested workloads, it is slightly more consistent relative to its peers. For users running legacy DirectX 9 or OpenGL applications, the GTX 950M’s 16 ROPs and 40 TMUs might suffice, but the GTX 1660 Ti’s 48 ROPs and 96 TMUs provide far higher fill rates.

In practical terms, the GTX 1660 Ti wins for any modern gaming or professional workload that demands high frame rates or fast compute. The GTX 950M wins only in the niche scenario where a Vulkan application is poorly optimized for Turing hardware, or where the lower power draw (75W vs 120W) is a critical constraint.

Specification Differences

The specification sheets reveal stark contrasts across every major component. The GTX 950M features 640 shading units, 40 TMUs, and 16 ROPs, while the GTX 1660 Ti boasts 1,536 shading units, 96 TMUs, and 48 ROPs—a 2.4x increase in shading units and a 3x increase in both TMUs and ROPs. Memory configurations diverge sharply: the GTX 950M uses 4GB of DDR3 on a 128-bit bus, yielding 28.80 GB/s bandwidth, whereas the GTX 1660 Ti uses 6GB of GDDR6 on a 192-bit bus, delivering 288.0 GB/s—exactly 10 times the bandwidth.

Clock speeds also differ, with the GTX 950M running at a base of 993 MHz and boost of 1124 MHz, compared to the GTX 1660 Ti’s base of 1500 MHz and boost of 1770 MHz. The memory clocks are 900 MHz (1800 Mbps effective) for the GTX 950M versus 1500 MHz (12 Gbps effective) for the GTX 1660 Ti. Pixel rate jumps from 17.98 GPixel/s to 84.96 GPixel/s, and texture rate from 44.96 GTexel/s to 169.9 GTexel/s. FP32 performance is a massive leap: 1,438.7 GFLOPS versus 5.437 TFLOPS. The GTX 1660 Ti also supports FP16 at 10.87 TFLOPS (2:1), which the GTX 950M does not list.

Power and physical characteristics differ as well. The GTX 950M is an IGP with no power connectors and a 75W TDP, while the GTX 1660 Ti is a dual-slot card requiring a 1x 8-pin connector and a 300W suggested PSU. The bus interface is PCIe 3.0 x8 for the GTX 950M versus PCIe 3.0 x16 for the GTX 1660 Ti. Display outputs vary: the GTX 950M is "Portable Device Dependent," while the GTX 1660 Ti offers 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a.

Architecture Differences

The architectural gulf is generational. The GTX 950M uses the GM107 chip built on TSMC’s 28nm process, employing the Maxwell architecture. It packs 1,870 million transistors on a 148 mm² die, yielding a density of 12.6M per mm². The GTX 1660 Ti, in contrast, uses the TU116 chip on a 12nm process, implementing the Turing architecture. It contains 6,600 million transistors on a 284 mm² die, achieving 23.2M per mm²—a density increase of roughly 84%.

The GTX 950M belongs to the GeForce 900M generation and supports DirectX 12 (11_0), while the GTX 1660 Ti is part of the GeForce 16 series and supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4, but the feature set differs. The GTX 1660 Ti’s FP16 capability (10.87 TFLOPS at 2:1 ratio) indicates a more modern compute pipeline, likely with concurrent execution of FP32 and FP16 operations. Neither GPU features ray tracing cores or tensor cores, as both are listed as null in the data.

The GTX 1660 Ti’s predecessor is listed as "GeForce 10," and its successor is "GeForce 20," placing it as a transitional Turing product without the RTX-specific hardware. The GTX 950M’s predecessor is GeForce 800M, with a successor of GeForce 10 Mobile. The GTX 1660 Ti was released on 2019-02-21, while the GTX 950M predates it by nearly four years, releasing on 2015-03-12. Both are end-of-life, but the GTX 1660 Ti’s launch MSRP was 279 USD.

The Verdict

From the data, the NVIDIA GeForce GTX 1660 Ti is the superior choice for nearly all performance-sensitive applications. Its OpenCL score is 5.2 times higher, its memory bandwidth is 10 times greater, and its FP32 throughput is nearly 4 times higher. The GTX 1660 Ti’s 6GB of GDDR6 memory versus the GTX 950M’s 4GB of DDR3 is not just a capacity increase—it is a fundamental bandwidth advantage that impacts every texture-heavy workload. For gaming at modern resolutions or running compute tasks, the GTX 1660 Ti is the clear pick.

The GTX 950M, however, has its own rationale. It wins the Vulkan test with a 12% margin, which could matter for users running specific Vulkan-based titles. Its average benchmark score (8,135) is higher than the GTX 1660 Ti’s (7,723), and its 42nd percentile ranking slightly edges out the 41st percentile of the newer card. The GTX 950M’s 75W TDP and IGP form factor make it suitable for thin-and-light laptops where power draw and space are at a premium. If a user prioritizes Vulkan performance or needs a low-power integrated solution, the GTX 950M remains viable.

The GTX 1660 Ti’s launch MSRP of 279 USD, stated once here, reflects its positioning as a mainstream desktop card, whereas the GTX 950M was designed for mobile integration. The specification sheet shows the GTX 1660 Ti is a dual-slot card measuring 229 mm in length, 111 mm in height, and 35 mm in width, with a 300W suggested PSU. The GTX 950M has no such requirements. For a desktop builder, the GTX 1660 Ti is the obvious choice. For a legacy laptop user, the GTX 950M’s Vulkan win and higher aggregate average make it a defensible holdover. Ultimately, the data favors the GTX 1660 Ti for raw performance, but the GTX 950M’s niche wins prevent a total shutdown.